Abstract
The Cognitive Performance Scale (CPS) in the widely used interRAI suite of instruments is of interest to clinicians and policy makers as a potential screening mechanism for detecting dementia. However, there has been little evaluation of the CPS in home care settings. This retrospective diagnostic study included 134 older adults (age ≥ 65) who were discharged from two acute psychogeriatric inpatient units or assessed in two memory clinics. The reference test was a diagnosis of clinical dementia, and the index test was interRAI CPS measured within 90 days of discharge. The overall accuracy of the CPS was good, with an area under the Receiver Operating Characteristic curve of 0.82 (95% CI = 0.75-0.89). The optimal cut point was 1/2, coinciding with the recommended cut point, with good sensitivity (0.90, 95% CI = 0.81-0.96) but poor specificity (0.60, 95% CI = 0.46-0.72). Positive predictive value improved from 0.72 (95% CI = 0.66-0.78) to 0.89 (95% CI = 0.75-0.96) when using a cut point of 2/3 instead of 1/2. If the results of the present study are replicated with more generalisable interRAI samples, older adults with a CPS of 3 or above, but without a formal diagnosis of dementia, should be referred for further cognitive assessment.